rulm

1
{
2
    "trainer": {
3
        "evaluation_strategy": "steps",
4
        "per_device_train_batch_size": 2,
5
        "per_device_eval_batch_size": 2,
6
        "gradient_accumulation_steps": 64,
7
        "eval_steps": 50,
8
        "save_steps": 50,
9
        "logging_steps": 5,
10
        "learning_rate": 0.00025,
11
        "num_train_epochs": 4,
12
        "lr_scheduler_type": "cosine",
13
        "warmup_steps": 30,
14
        "fp16": false,
15
        "bf16": true,
16
        "torch_compile": false,
17
        "optim": "adamw_torch"
18
    },
19
    "lora": {
20
        "r": 16,
21
        "lora_alpha": 16,
22
        "lora_dropout": 0.05,
23
        "bias": "none",
24
        "target_modules": ["q_proj", "v_proj", "k_proj", "o_proj"],
25
        "task_type": "CAUSAL_LM"
26
    },
27
    "use_flash_attention_2": true,
28
    "load_in_8bit": false,
29
    "load_in_4bit": true,
30
    "only_target_loss": true,
31
    "mode": "chat",
32
    "templates_path": "internal_prompts/saiga_v2.json",
33
    "model_name": "models/Yarn-Mistral-7b-128k",
34
    "model_type": "causal",
35
    "max_tokens_count": 2000
36
}
37

38